An EM-based Semi-Parametric Mixture Model Approach to the Regression Analysis of Competing-Risks Data

We consider a mixture model approach to the regression analysis of competing-risks data. Attention is focused on inference concerning the effects of factors on both the probability of occurrence and the hazard rate conditional on each of the failure types. These two quantities are specified in the mixture model using the logistic model and the proportional hazards model, respectively. We propose a semi-parametric mixture method to estimate the logistic and regression coefficients jointly, whereby the component-baseline hazard functions are completely unspecified. Estimation is based on maximum likelihood on the basis of the full likelihood, implemented via an expectation-conditional maximization (ECM) algorithm. Simulation studies are performed to compare the performance of the proposed semi-parametric method with a fully parametric mixture approach. The results show that when the component-baseline hazard is monotonic increasing, the semi-parametric and fully parametric mixture approaches are comparable for mildly and moderately censored samples. When the component-baseline hazard is not monotonic increasing, the semi-parametric method consistently provides less biased estimates than a fully parametric approach and is comparable in efficiency in the estimation of the parameters for all levels of censoring. The methods are illustrated using a real data set of prostate cancer patients treated with different dosages of the drug diethylstilbestrol. Copyright (C) 2003 John Wiley Sons, Ltd.

Rab23, a negative regulator of hedgehog signaling, localizes to the plasma membrane and the endocytic pathway

The regulation of hedgehog signaling by vesicular trafficking was exemplified by the finding that Rab23, a Rab-GTPase vesicular transport protein, is mutated in open brain mice. In this study, the localization of Rab23 was analyzed by light and immunoelectron microscopy after expression of wild-type (Rab23-GFP), constitutively active Rab23 (Rab23Q68L-GFP), and inactive Rab23 (Rab23S23N-GFP) in a range of mammalian cell types. Rab23-GFP and Rab23Q68L-GFP were predominantly localized to the plasma membrane but were also associated with intracellular vesicular structures, whereas Rab23S23N-GFP was predominantly cytosolic. Vesicular Rab23-GFP colocalized with Rab5Q79L and internalized transferrin-biotin, but not with a marker of the late endosome or the Golgi complex. To investigate Rab23 with respect to members of the hedgehog signaling pathway, Rab23-GFP was coexpressed with either patched or smoothened. Patched colocalized with intracellular Rab23-GFP but smoothened did not. Analysis of patched distribution by light and immunoelectron microscopy revealed it is primarily localized to endosomal elements, including transferrin receptor-positive early endosomes and putative endosome carrier vesicles and, to a lesser extent, with LBPA-positive late endosomes, but was excluded from the plasma membrane. Neither patched or smoothened distribution was altered in the presence of wild-type nor mutant Rab23-GFP, suggesting that despite the endosomal colocalization of Rab23 and patched, it is likely that Rab23 acts more distally in regulating hedgehog signaling.

The use of 16s rDNA restriction fragment length polymorphism analysis for the identification of non-fermenting gram negative bacilli in a cystic fibrosis microbiology referral service

The utility of 16s rDNA restriction fragment length polymorphism (RFLP) analysis for the partial genomovar differentiation of Burkholderia cepacia complex bacterium is well documented. We compared the 16s rDNA RFLP signatures for a number of non-fermenting gram negative bacilli (NF GNB) LMG control strains and clinical isolates pertaining to the genera Burkholderia, Pseudomonas, Achromobacter (Alcaligenes), Ralstonia, Stenotrophomonas and Pandoraea. A collection of 24 control strain (LMG) and 25 clinical isolates were included in the study. Using conventional PCR, a 1.2 kbp 16s rDNA fragment was generated for each organism. Following restriction digestion and electrophoresis, each clinical isolate RFLP signature was compared to those of the control strain panel. Nineteen different RFLP signatures were detected from the 28 control strains included in the study. TwentyoneyTwenty- five of the clinical isolates could be classified by RFLP analysis into a single genus and species when compared to the patterns produced by the control strain panel. Four clinical B. pseudomallei isolates produced RFLP signatures which were indistinguishable from B. cepacia genomovars I, III and VIII. The identity of these four isolates were confirmed using B. pseudomallei specific PCR. 16s rDNA RFLP analysis can be a useful identification strategy when applied to NF GNB, particularly for those which exhibit colistin sulfate resistance. The use of this molecular based methodology has proved very useful in the setting of a CF referral laboratory particularly when utilised in conjunction with B. cepacia complex and genomovar specific PCR techniques. Species specific PCR or sequence analysis should be considered for selected isolates; especially where discrepancies between epidemiology, phenotypic and genotypic characteristics occur.

The genetics of glycosylation in Gram-negative bacteria

In recent years there has been a dramatic increase in reports of glycosylation of proteins in various Gram-negative systems including Neisseria meningitidis, Neisseria gonorrhoeae, Campylobacter jejuni, Pseudomonas aeruginosa, Escherichia coli, Caulobacter crescentus, Aeromonas caviae and Helicobacter pylori. Although this growing list contains many important pathogens (reviewed by Benz and Schmidt [Mol. Microbiol. 45 (2002) 267-276]) and the glycosylations are found on proteins important in pathogenesis such as pili, adhesins and flagella the precise role(s) of the glycosylation of these proteins remains to be determined. Furthermore, the details of the glycosylation biosynthetic process have not been determined in any of these systems. The definition of the precise role of glycosylation and the mechanism of biosynthesis will be facilitated by a detailed understanding of the genes involved. (C) 2002 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.

Style and performance of international socially responsible funds in Europe

This paper analyses the performance and investment styles of internationally oriented Socially Responsible Investment (SRI)funds, domiciled in eight European markets, in comparison with characteristics-matched conventional funds. To the best of our knowledge, this is the first multi-country study, focused on international SRI funds (investing in Global and in European equities), to combine the matched-pairs approach with the use of robust conditional multi-factor performance evaluation models, which allow for both time-varying alphas and betas and also control for home biases and spurious regression biases.In general, the results show that differences in the performance of international SRI funds and their conventional peers are not statistically significant. Regarding investment styles, SRI and conventional funds exhibit similar factor exposures in most cases. In addition,conventional benchmarks present a higher explaining power of SRI fund returns than SRI benchmarks. Our results also show significant differences in the investment styles of SRI funds according to whether they use “best-in-class” screening strategies or not. When compared to SRI funds that employ simple negative and/or positive screens, SRI “best-in-class” funds present significantly lower exposures to small caps and momentum strategies and significantly higher exposures to local stocks.

ABOUT PSYCHOLOGICAL VARIABLES IN APPLICATION SCORING MODELS

The purpose of this study is to investigate the contribution of psychological variables and scales suggested by Economic Psychology in predicting individuals’ default. Therefore, a sample of 555 individuals completed a self-completion questionnaire, which was composed of psychological variables and scales. By adopting the methodology of the logistic regression, the following psychological and behavioral characteristics were found associated with the group of individuals in default: a) negative dimensions related to money (suffering, inequality and conflict); b) high scores on the self-efficacy scale, probably indicating a greater degree of optimism and over-confidence; c) buyers classified as compulsive; d) individuals who consider it necessary to give gifts to children and friends on special dates, even though many people consider this a luxury; e) problems of self-control identified by individuals who drink an average of more than four glasses of alcoholic beverage a day.

DETERMINANTS OF THE SUCCESS OF GLOBAL AND LOCAL BRANDS IN LATIN AMERICA

ABSTRACTThe purpose of this paper is to address the issue of the implementation of global and local brands in Latin America by drawing on contingency theory to develop and test hypotheses relating to how product category characteristics affect the success of global and local brands in the region. Hypotheses are tested using data obtained from top brands rankings reported in five Latin American markets (Argentina, Brazil, the Caribbean and Central America, Chile and Mexico). The study design considers estimating a logistic regression on a binomial dependent variable measuring whether 475 top brands are global or local brands, with product category characteristics as independent variables. Results reveal that product categories related to subscriptions, local tastes, high-tech, and global citizenship do have an impact on the success of global and local brands in Latin America.

Developmental and evolutionary assumptions in a study about the impact of premature birth and low income on mother–infant interaction

In order to study the impact of premature birth and low income on mother–infant interaction, four Portuguese samples were gathered: full-term, middle-class (n=99); premature, middle-class (n=63); full-term, low income (n=22); and premature, low income (n=21). Infants were filmed in a free play situation with their mothers, and the results were scored using the CARE Index. By means of multinomial regression analysis, social economic status (SES) was found to be the best predictor of maternal sensitivity and infant cooperative behavior within a set of medical and social factors. Contrary to the expectations of the cumulative risk perspective, two factors of risk (premature birth together with low SES) were as negative for mother–infant interaction as low SES solely. In this study, as previous studies have shown, maternal sensitivity and infant cooperative behavior were highly correlated, as was maternal control with infant compliance. Our results further indicate that, when maternal lack of responsiveness is high, the infant displays passive behavior, whereas when the maternal lack of responsiveness is medium, the infant displays difficult behavior. Indeed, our findings suggest that, in these cases, the link between types of maternal and infant interactive behavior is more dependent on the degree of maternal lack of responsiveness than it is on birth status or SES. The results will be discussed under a developmental and evolutionary reasoning

Reproductive outcomes in an area adjacent to a petrochemical plant in southern Brazil

OBJECTIVE: To evaluate possible adverse reproductive outcomes in an area adjacent to a petrochemical plant in southern Brazil. METHODS: A review of 17,113 birth records of the main hospital of the municipality of Montenegro, southern Brazil, from 1983 to 1998 was carried out. Three groups of cases were selected: (1) newborns with major congenital malformations; (2) newborns with low birth weight (<2,500 g); and (3) stillborns (>500 g). A control was assigned to each case. Controls were the first newborns weighing > or = 2,500 g without malformations and of case-matching sex. Mother's residence during pregnancy was used as an exposure parameter. Statistical analyses were performed using Chi-square test or Fisher test, odds ratio, 0.05 significance level, and 95% confidence interval. RESULTS: For unadjusted analysis, it was found a correlation between low birth weight and geographical proximity of mother's residence to the petrochemical plant (OR = 1.66; 95% CI = 1.01--2.72) or residence on the way of preferential wind direction (OR = 1.62; 95% CI = 1.03--2.56). When other covariates were added in the conditional logistic regression (maternal smoking habits, chronic disease and age), there was no association. CONCLUSIONS: Despite final results were negative, low birth weight could be a good parameter of environmental contamination and should be closely monitored in the studied area.

Long-term price range forecast applied to risk management using regression models

Long-term contractual decisions are the basis of an efficient risk management. However those types of decisions have to be supported with a robust price forecast methodology. This paper reports a different approach for long-term price forecast which tries to give answers to that need. Making use of regression models, the proposed methodology has as main objective to find the maximum and a minimum Market Clearing Price (MCP) for a specific programming period, and with a desired confidence level α. Due to the problem complexity, the meta-heuristic Particle Swarm Optimization (PSO) was used to find the best regression parameters and the results compared with the obtained by using a Genetic Algorithm (GA). To validate these models, results from realistic data are presented and discussed in detail.

Job control, job demands, social support at work and health among adolescent workers

OBJECTIVE: To evaluate physical and psychological dimensions of adolescent labor (such as job demands, job control, and social support in the work environment), and their relation to reported body pain, work injuries, sleep duration and daily working hours. METHODS: A total of 354 adolescents attending evening classes at a public school in São Paulo, Brazil, answered questionnaires regarding their living and working conditions (Karasek's Job Content Questionnaire, 1998), and their health status. Data collection took place in April and May 2001. Multiple logistic regression analysis was used to determine relations among variables. RESULTS: Psychological job demands were related to body pain (OR=3.3), higher risk of work injuries (OR=3.0) and reduced sleep duration in weekdays (Monday to Thursday) (p<0.01). Lower decision authority in the workplace (p=0.03) and higher job security (p=0.02) were related to longer daily working hours. CONCLUSIONS: It was concluded that besides physical stressors, psychological factors are to be taken into account when studying adolescent working conditions, as they may be associated with negative job conditions and health effects.

Factors associated to smoking habit among older adults (The Bambuí Health and Aging Study)

OBJECTIVE: To describe the characteristics and associated factors of the smoking habit among older adults. METHODS: A population-based study was carried out comprising 1,606 (92.2%) older adults (>60 years old) living in the Bambuí town, Southeastern Brazil in 1997. Data was obtained by means of interview and socio-demographic factors, health status, physical functioning, use of healthcare services and medication were considered. The multiple multinomial logistic regression was used to assess independent associations between smoking habits (current and former smokers) and the exploratory variables. RESULTS: The prevalence of current and past smoking was 31.4% and 40.2% among men, and 10.3% and 11.2% among women, respectively (p<0.001). Among current smokers, men consumed a larger number of cigarettes per day and started the habit earlier than women. Among men, current smoking presented independent and negative association with age (>80 years) and schooling (>8 years) and positive association with poor health perception and not being married. Among women, independent and negative associations with current smoking were observed for age (75-79 and >80 years) and schooling (4-7 and >8 years). CONCLUSIONS: Smoking was a public health concern among older adults in the studied community, particularly for men. Yet, in a low schooling population, a slightly higher level was a protective factor against smoking for both men and women. Programs for reducing smoking in the elderly population should take these findings into consideration.

Negative modulation of human osteoclastogenesis by antiepileptic drugs

Bone is constantly being molded and shaped by the action of osteoclasts and osteoblasts. A proper equilibrium between both cell types metabolic activities is required to ensure an adequate skeletal tissue structure, and it involves resorption of old bone and formation of new bone tissue. It is reported that treatment with antiepileptic drugs (AEDs) can elicit alterations in skeletal structure, in particular in bone mineral density. Nevertheless, the knowledge regarding the effects of AEDs on bone cells are still scarce, particularly on osteoclastic behaviour. In this context, the aim of this study was to investigate the effects of five different AEDs on human osteoclastic cells. Osteoclastic cell cultures were established from precursor cells isolated from human peripheral blood, and were maintained in the absence (control) or in the presence of 10-8-10-4 M of different AEDs (valproate, carbamazepine, gabapentin, lamotrigine and topiramate). Cell cultures were characterized throughout a 21-day period for tartrate-resistant acid phosphatase (TRAP) activity, number of TRAP+ multinucleated cells, presence of cells with actin rings and expressing vitronectin and calcitonin receptors, and apoptosis rate. Also, the involvement of several signaling pathways on the cellular response was addressed. All the tested drugs were able to affect osteoclastic cell development, although with different profiles on their osteoclastogenic modulation properties. Globally, the tendency was to inhibit the process. Furthermore, the signaling pathways involved in the process also seemed to be differentially affected by the AEDs, suggesting that the different drugs may affect osteoclastogenesis through different mechanisms. In conclusion, the present study showed that the different AEDs had the ability to negatively modulate the osteoclastogenesis process, shedding new light towards a better understanding of how these drugs can affect bone tissue.

Obesity and gestational weight gain: cesarean delivery and labor complications

OBJECTIVE: To assess the association between pre-gestational obesity and weight gain with cesarean delivery and labor complications. METHODS: A total of 4,486 women 20-28 weeks pregnant attending general prenatal care clinics of the national health system in Brazil from 1991 to 1995 were enrolled and followed up through birth. Body mass index categories based on prepregnancy weight and total weight gain were calculated. Associations between body mass index categories and labor complications were adjusted through logistic regression analysis. RESULTS: Obesity was present in 308 (6.9%) patients. Cesarean delivery was performed in 164 (53.2%) obese, 407 (43.1%) pre-obese, 1,045 (35.1%) normal weight and 64 (24.5%) underweight women. The relative risk for cesarean delivery in obese women was 1.8 (95% CI: 1.5-2.0) compared to normal weight women. Greater weight gain was particularly associated with cesarean among the obese (RR 4th vs 2nd weight gain quartile 2.2; 95% CI: 1.4-3.2). Increased weight at the beginning of pregnancy was associated with a significantly higher adjusted risk of meconium with vaginal delivery and perinatal death and infection in women submitted to cesarean section. Similarly, greater weight gain during pregnancy increased the risk for meconium and hemorrhage in women submitted to vaginal delivery and for prematurity with cesarean. CONCLUSIONS: Pre-gestational obesity and greater weight gain independently increase the risk of cesarean delivery, as well as of several adverse outcomes with vaginal delivery. These findings provide further evidence of the negative effects of prepregnancy obesity and greater gestational weight gain on pregnancy outcomes.

Motivations which influence volunteers' satisfaction

In this paper we attempt to find out which motivations influence volunteers satisfaction. So, in a first moment, we categorize satisfaction through a factor analysis and then use linear regression models to find out the relations previously mentioned. Furthermore we analyse the existence of correlations between some demographic attributes and the other variables. In this research the population under study includes volunteers that work in the health area, more precisely volunteers that work in hospital and have direct contact with patients and their families. We obtained a total of 327 questionnaires and after excluding incomplete answers we get 304 questionnaires which embody a response rate of 36%. The outcomes of our analysis reveal that we can categorize satisfaction into intrinsic and extrinsic categories and show that motivations related to belonging and protection and career recognition are the ones influencing extrinsic satisfaction; motivations associated to development and learning and altruism are the ones with higher effect in intrinsic satisfaction. There are some negative correlations between age and extrinsic satisfaction, between years of participation and extrinsic satisfaction, between education and the motivation related to belonging and protection and between age and the motivation related to career recognition. There is only one positive correlation between hours per week and intrinsic satisfaction. The results offer new insights for research about volunteers’ motivations, motive fulfillment and volunteer satisfaction. Furthermore the outcomes advocate that to ensure satisfied volunteers, their motivations must be identified in a timely and appropriate moment, meaning that should happened as early as possible.